Modeling the MOC

Ronald J StoufferGeophysical Fluid Dynamics Laboratory

NOAAThe views described here are solely those of the presenter and not of GFDL/NOAA/DOC or any other agency or institution.

http://www.andrill.org/iceberg/blogs/julian/images/greatoceanconveyor.jpg

Modeling the MOCOutline

• Role of MOC in maintaining mean climate– Northward Heat Transport – Northward Salt transport

• Role of MOC in Abrupt Climate Change– Unforced– Forced

• Predictability

How well do models simulate the T, S structure in the ocean?

AR4 ensemble mean error Temperature

IPCC WGI Chapter 8

-2.5 +2.5

AR4 ensemble mean errorSalinity

90N 90S

-0.2 0.2-0.4 0.4

IPCC WGI Chapter 8 Suppl. Material

EQ

PSU-1.0 +1.0

Role of MOC in Heat Transport

Impact of MOC on Climate(SAT)

AOGCMs

EMICs

MOC “on” minus “off” oC

Conclusion – MOC warms NH, locally large values, MOC cools SH

Impact of MOC on Climate

Zonally averaged Precipitation differences(mm/day)

Red lines controlBlue lines difference

Dashed – EMICsSolid - AOGCMs

Conclusion – MOC on => ITCZ toward north

Impact of MOC on ClimateSalinity MOC “on” minus “off”

(PSU)

Role of MOC in maintaining mean climate

• North Atlantic saltier than without MOC– Rest of world ocean surface more fresh

• Northern Hemisphere warmer than without– Particularly the N Atlantic– Role of atmosphere mixing heat

Ocean MOC Role in Abrupt Climate Changes

• Unforced – Hall and Stouffer Nature

• Forced– Idealized (Hosing ) - Stouffer et al. JoC 2006

TransientAn Anomalous Event (Unforced)

Maximum Negative Anomaly

Maximum Positive Anomaly

Surface Air TemperatureDecadal Mean Difference

Surface variables/THC

Summary of Physical

Mechanism

Unforced MOC variability

• If model if “realistic”– Can we predict this event?– Complicating GHG changes– Possible explanation of some parts of paleo-

record

Experimental DesignManabe Climate Model “MCM”

• R30 AOGCM coupled model

• Idealized Water hosing– 1 Sv for 100 years– After 100 years, stop hosing - allow recovery

• Case 1: Hosing 50N to 70N in Atlantic

• Case 2: Hosing south of 60S in Southern Ocean

• Compare to long control integration

Simulated Global MOC

SH Index box NH Index box

Atlantic THC Response

SV

Years

Atlantic THCdoes notrespond in aseesaw-likemanner

NH Hosing –NH THCshuts down

SH THC Response

SV

Years

SH Hosing –

SH THCweakens.

SH THCdoes notshut down

SAT Difference mapSH hosing

K

Years 51-100 hosing minus 1-200 control

Surface Salinity Response

0 100 200 Years

0 100 200 Years

NH

SH

NH SSS anomaly –Intense and confined

SH SSS anomaly –Weaker and spreads

PSU

Differences in Sea Surface Salinity (PSU)

Southern Freshwater Escape

25 years 100 years

Hosing minus Control

Sea Surface Temperature Response

0 100 200 Years

0 100 200 Years

NH

SH

Response more symmetrical than SSSMagnitude also becoming more similar

K

Surface Air Temperature Response

0 100 200 Years

0 100 200 Years

NH

SH

Response remarkably symmetrical (first 100 yrs)Magnitude very similar

K

Precipitation Response

0 100 200 Years

0 100 200 Years

NH

SH

Response very symmetricMagnitude very similarITCZ shifts toward warmer hemisphere

Cm/day

Hosing Experiment Summary

• Symmetrical Atmospheric Response

• Much less symmetry in ocean

• Why?– Strong Circum-Antarctica winds– Northward flowing surface waters– Freshwater “escapes” into other basins

• Far a field impacts• Less local impacts

MOC Predictability

• Unforced Changes– Are MOC predictable?– How long into future– Manifest in surface changes?

• Forced - GHG increase – Does MOC weaken?– How much?– Likelihood of complete shutdown?

Predictability Investigations just starting

• Perfect model experiments

• Use ICs from long control– Ocean ICs unchanged from control– Atmosphere ICs shifted in time by day or 2

• Probably “best case” for predictability– No model errors– Ocean ICs perfectly known

Predictability ofAtlantic MeridionalOverturning Circulation(AMOC) in GFDLCM2.1 Climate Model

Are past ocean observations good enough to constrain MOC?

• Past ocean observations mainly XBTs– Temperature only– Upper 700 m or so

• Since 2003 or so – ARGO– T, S– Upper 2 km

Can Observations constrain the MOC?

Need ARGOand atmosphericdata to constrain MOC

Other research suggests this may be too pessimistic.

A

AR4 WG1 Assessment: MOC very likely to weaken MOC shutdown very unlikely

MOC and Forced Climate Change

Why does MOC slow down as GHG increase?

• Role of surface fluxes– Heat fluxes– Water fluxes

Design of Partially Coupled Experiments

Gregory et al. 2006 GRL

• Run control and 1% per year CO2 increase experiment– Save out surface fluxes

• Use water fluxes from control in 1% run– TRAD_CH2O

• Use water fluxes from 1% in control– CRAD_TH2O

• Isolates role of heat and water fluxes

Summary of Partially Coupled Experiment

Conclusions:

1. Heat fluxes changes always weaken MOC

2. Water fluxes changes mixed, but usually weaken MOC

3. Response to heat flux changes fairly uniform

Warming greatest over land and at most high northern latitudes and least over Southern

Ocean and parts of the North Atlantic Ocean

• Weakening of MOC contributes to minimum in cooling in N Atlantic => smaller climate change => a positive impact?

Surface Warming Pattern

A1B, 2090-2099 relative to 1980-1999

IPCC WGI SPM

Summary:Ocean’s Role in Abrupt Climate

Change• Unforced

– Possible to have large abrupt climate changes in AOGCMs

• Forced– Idealized

• Allows easy study of climate response• Application to paleo-data and future climate changes

– GHG increase

• Predictability of MOC changes

Questions

• Basic Issues– Why does the MOC exist?– How much mixing is there in the ocean?

• Do models have too much/too little? • Impacts of mixing on the MOC• What are the physical processes?

– What is the role of ocean eddies?– How does MOC changes impact biology and

associated changes in atmospheric pCO2?

Questions

• Variability– Are observations good enough to constrain MOC? – Are the MOC changes predictable?

• Time scale?• Does it matter for where people live?• Paleo-data tests?

– What is role of MOC variability/changes in tropical Atlantic SST/hurricane changes?

• Future– Is future weakening of the MOC “bad”?– Interactions with Greenland/Antarctic ice melting

Thank you